Photocatalysis is an effective and eco-friendly technology, which shows promise for applications in the field of environmental purification and solar energy conversion [1-3]. At present, photocatalytic materials are primarily classified into three kinds: first, inorganic photocatalytic materials, such as TiO2 [4], WO3 [5], Fe2O3 [6], Ga2O3 [7], Bi2O2CO3 [8], MoS2 [9], Ag3PO4 [10], and Ta2O5 [11]; second, organic photocatalytic materials, such as graphite carbon nitride [12], polyimide [13], and conductive poly-1, 4-diphenylbutadiene [14]; third, elementary photocatalytic materials [15, 16], such as Bi, Si, S, P, and Se. However, most of the above-mentioned photocatalytic materials are limited by their low utilization of solar energy and high recombination rates of photo-generated carriers. Therefore, there is an urgent need to develop cost-efficient and simple methods to prepare novel visible-light driven photocatalytic materials with high and stable photoactivity.
In recent years, BiOBr has attracted considerable attention owing to its appropriate band gap position for photocatalytic processes and unique layered nanostructure. However, BiOBr also suffers from low efficiency in terms of its visible light utilization and has high recombination rates of photo-generated electron-hole pairs, which prevent practical application. To date, it has been reported that various strategies can be used to enhance the photocatalytic performance of BiOBr, such as architecture control, metal or non-metallic doping, plasma decorating, and heterojunction construction [17, 18]. Density functional theory (DFT) calculation shows that the valence band (VB) of BiOBr is composed of hybrid orbitals of Br 4p and O 2p, and the conduction band (CB) consist of Bi 6p orbitals. More recently, it has been found that a lower Br/O ratio lowers the semiconductor valence band position, which decreases the band gap and increases utilization of visible light [19]. Hence, an important direction in future research is the exploration, preparation, and application of new types of BixOyBrz with various stoichiometric compositions, such as Bi5O7Br, Bi3O4Br, Bi12O17Br2, Bi24O31Br10, and Bi4O5Br2. Moreover, BixOyBrz with different compositions can be produced by thermodynamic Br-substitution reactions. Xia et al. [20] reported that Bi4O5Br2ultrathin nanosheets can be synthesized by a reactive ionic liquid-assisted solvothermal route. These materials showed good photocatalytic performance for removal of the colorless antibiotic agent ciprofloxacin under visible light irradiation. Xu et al. [21] prepared Bi12O17Br2/Bi24O31Br10 hierarchical heterostructures by calcining BiOBr/Bi(OHC2O4)·2H2O precursors at 400 ℃ for 2 h in air. The as-obtained Bi12O17Br2/Bi24O31Br10 samples exhibited excellent visible light photocatalytic performance for the degradation of phenol and rhodamine B. However, the preparation process was complicated and restrictive, featuring high temperatures and pressures. These features of the synthesis added to the cost and processing time, greatly limiting the large-scale application of BixOyBrz photocatalytic materials. Nitrogen oxides, as major air contaminants, have serious effects on human health and the environment. Furthermore, the conversion pathway of nitrogen oxide by visible light photocatalytic oxidation over BixOyBrz materials has not yet been reported. These materials could have an important role in the field of environmental purification. Therefore, it is necessary to explore a simple and environment-friendly method to prepare efficient and stable BixOyBrz photocatalytic materials, and to investigate their physicochemical properties and mechanism of visible light photocatalytic nitrogen oxide oxidation.
In this study, Bi12O17Br2 and Bi4O5Br2 nanosheets were fabricated by hydrothermal and room-temperature deposition methods, respectively. Bi4O5Br2exhibited higher photocatalytic performance than that of Bi12O17Br2 in removal of NO under visible light irradiation. In addition, on the basis of in situ diffuse reflectance infrared Fourier transform spectroscopy (DRIFTS) investigations, a reaction mechanism for the photocatalytic NOx removal was proposed. The present work could provide a new route to the synthesis and optimization of BixOyBrz photocatalysts with various stoichiometric compositions for environmental purification and solar energy conversion.
All chemical reagents used in this experiment were of analytical purity. A typical preparation process of the Bi12O17Br2 nanosheet is as follows. 1.33 g BiBr3 and 50 mL absolute alcohol were added to a Teflon-lined stainless-steel autoclave, and vigorously stirred for 30 min. Subsequently, 12.6 mL of NaOH solution (2 mol/L) was added dropwise into the BiBr3 solution and stirred for 4 h. The obtained precursor suspension was then subjected to hydrothermal treatment at 150 ℃ for 24 h. After the reaction was complete, the obtained solid sample was washed twice with deionized water and absolute ethanol, and then dried at 60 ℃ to obtain the Bi12O17Br2 sample. The Bi4O5Br2 nanosheets were formed under the same conditions without the hydrothermal treatment.
The crystal structure of the sample was analyzed by X-ray diffraction (XRD: model D/max RA, Rigaku Co., Japan). The morphology of the samples was analyzed with a scanning electron microscope (SEM: JEOL model JSM-6490, Japan) and transmission electron microscope (TEM: JEM-2010, Japan). The specific surface area (BET) and pore structure of the samples were determined by N2 adsorption-analyzer (ASAP 2020, USA). The optical properties of the samples were analyzed by UV-Vis diffuse spectroscopy (UV-vis DRS: UV2550PC, Shimadzu, Japan). The main active free radical species in the photocatalytic reactions were analyzed by electron spin resonance (ESR: FLsp920, UK). Oxygen vacancies in the samples were analyzed by electron paramagnetic resonance spectroscopy (EPR: Bruker ESP 500, Germany).
The photocatalytic activity was evaluated by monitoring the concentration of NO in the continuous stream reactor at room temperature. 0.2 g portion of the sample was ultrasonically dispersed on a 12-mm diameter glass disk, and dried at 70 ℃. After natural cooling, the mixture was placed in a rectangular reactor with a capacity of 4.5 L (30 cm × 15 cm × 10 cm). The top was covered with a layer of quartz glass, and a 150-W commercial tungsten lamp was placed 20 cm above the reactor vertically. For visible light photocatalytic activity tests, filters were used to remove wavelengths below 420 nm (i.e., ultraviolet light) from the beam. A low concentration mixture of NO was prepared with standard air and a NO standard gas with an initial concentration of 100 ppm. The air flow rate was 2.4 L/min and the NO gas flow rate was 15 mL/min. The air flow and the NO gas stream were mixed via a three-way valve to obtain NO, the initial concentration of NO was diluted to about 600 ppb (balanced gas). The desired relative humidity level of the NO flow was controlled at 50% by passing the zero air streams through a humidification chamber. After adsorption-desorption equilibrium was achieved, the lamp was turned on. The NOx analyzer (Thermo Scientific, 42i-TL) measurements were performed at 1-min intervals and the concentrations of NO, NO2, and NOx (NOx including NO and NO2) were recorded. The removal rate (η) of NO was calculated as follows:
where C represents the NO concentration at the outlet of the reactor after the lamp was turned on; C0 represents the NO concentration when the adsorption-desorption equilibrium was reached before turning on the lamp.
The in situ DRIFTS instrument consisted of a Tensor Ⅱ FT-IR spectrometer (Bruker) equipped with an in situ reflection reaction chamber (Harrick), and the photocatalyst was placed in the reaction chamber. First, He gas (100 mL/min) was applied to remove residual carbohydrates, water, and carbon dioxide adsorbed on the surface of the photocatalyst. Second, real-time FT-IR spectra measured after ventilation with He gas were used as a background. Third, the reaction mixtures (50 mL/min NO and 50 mL/min O2) were introduced in the reaction chamber for 20 min. After the adsorption was completed, the sample was illuminated with a visible light source (MVL-210) for 60 min. The real time FT-IR spectra were detected at 2-min intervals. Finally, the light source was turned off. Infrared spectra were measured in the range between 600–4000 cm-1 and used to analyze the photocatalytic oxidation process of NO.
Fig. 1 shows the XRD patterns of Bi12O17Br2 and Bi4O5Br2 samples. All the diffraction peaks of Bi12O17Br2 correspond to the tetragonal structure of Bi12O17Br2(JCPDS Card No. 37-0701). All the diffraction peaks of Bi4O5Br2 could be indexed to the monoclinic structure of Bi4O5Br2(JCPDS Card No. 37-0699). Bi12O17Br2 and Bi4O5Br2featured strong diffraction peaks and no other peaks were found in the XRD patterns, which indicated that Bi12O17Br2 and Bi4O5Br2 had good crystallinity and high purity. In the alkaline environment, OH- gradually replaced Br- in BiOBr, resulting in the formation of Bi4O5Br2. Under the hydrothermal reaction conditions, OH- further replaced Br- in Bi4O5Br2, which contributed to the generation of Bi12O17Br2. Hence, Bi12O17Br2 was apparently obtained through reorganization of Bi4O5Br2. The chemical formation mechanism of Bi4O5Br2 and Bi12O17Br2 crystals was as follows: (1)–(4) [19-23].
Fig. 2(a) and (b) show SEM images of Bi12O17Br2. Bi12O17Br2, each demonstrating a compact layered structure formed by stacking of nanosheets of irregular shapes and sizes. As shown in Fig. 2(c), the Bi12O17Br2 nanosheets were relatively large and thick, and this closely packed layered structure is not conducive to formation of a porous structure. Fig. 2(d) shows a HRTEM image of Bi12O17Br2 nanosheets with a lattice spacing of 0.59 nm, which derived from the {006} plane of Bi12O17Br2.
Fig. 3(a) and (b) show SEM images of Bi4O5Br2. The Bi4O5Br2 formed as stacks of nanosheets and nanoparticles, which showed an irregular morphology. As shown in Fig. 3(c), the Bi4O5Br2 nanosheets and nanoparticles were deposited on each other to form loose structures, which might be beneficial in terms of increasing the specific surface area and pore structure of the material. Fig. 3(d) shows a HRTEM image of Bi4O5Br2 with a lattice spacing of 0.29 nm, attributed to the {113} plane of Bi4O5Br2.
Fig. 4 shows the N2 adsorption-desorption isotherms and corresponding pore size distribution curves of Bi12O17Br2 and Bi4O5Br2. As shown in Fig. 4(a), all isotherms could be classified as characteristic Ⅳ type isotherms (classified according to BDDT), which indicates that Bi12O17Br2 and Bi4O5Br2 had mesoporous structures [24]. The hysteresis loop shape was classified as H3 type, indicating the presence of slit-like pores formed by stacking of the nanosheets [25]. This result is consistent with the SEM and TEM observations. The pore size distribution curve (Fig. 4(b)) reveals that the Bi12O17Br2 sample had a large pore size of 85 nm and the pore sizes of Bi4O5Br2 were mainly concentrated at 2.5 and 65 nm. Moreover, the specific surface area and pore volume of Bi4O5Br2 (37.2 m2/g and 0.215 cm3/g) were considerably higher than those of Bi12O17Br2 (8.7 m2/g and 0.04 cm3/g).
As shown in Fig. 5, both Bi12O17Br2 and Bi4O5Br2 featured strong light absorption in the visible light region. Moreover, the absorption edges (λg) of Bi12O17Br2 and Bi4O5Br2 were approximately 430 and 468 nm. Thus, the band energies of Bi12O17Br2 and Bi4O5Br2 were respectively estimated to be 2.65 and 2.88 eV from the relationship Eg = 1240/λg.To further explore the effects of the band structure on the photocatalytic activities of Bi12O17Br2 and Bi4O5Br2, the corresponding CB and VB position of Bi12O17Br2 and Bi4O5Br2 were calculated. At the zero potential (pHzpc), the conduction band position (ECB) of a semiconductor can be calculated by as: ECB = X -EC-1/2Eg [26], where X is the absolute electronegativity of the semiconductor, EC is the energy of free electron (about 4.5 eV) with respect to the hydrogen potential level, and Eg is the band gap energy of the semiconductor. The CB and VB values of Bi12O17Br2 and Bi4O5Br2 are listed in Table 1.
To evaluate the intrinsic environmental contaminant purification capacity of Bi12O17Br2 and Bi4O5Br2, the photocatalytic removal of NO from air was examined. As shown in Fig. 6(a), the C/C0 for Bi12O17Br2 and Bi4O5Br2 decreased sharply owing to the photocatalytic removal at 5 min. However, the photocatalytic reaction products generated during the removal of NO might occupy active sites of the photocatalysts, resulting in a decrease of the photo-activity. After 30 min of visible light irradiation, the removal efficiencies of NO by Bi12O17Br2 and Bi4O5Br2 were 28.3% and 41.8%, respectively. Moreover, the removal efficiency of NO from the Bi4O5Br2 sample was 41.1% after five cycles of testing (Fig. 6(b)). This decrease might be attributed to blockage of reactive sites. However, the final removal efficiency after 30 min of irradiation did not show a large decrease compared with that of the removal rate test in Fig. 6(a), which indicates that Bi4O5Br2 had good photocatalytic stability.
To explore the main active radical species formed during the photocatalytic reaction, DMPO-•O2- and DMPO-•OH trapping experiments were performed on the Bi12O17Br2 and Bi4O5Br2 samples in methanol solution. As can be seen from Fig. 7, the DMPO-•OH signal can be clearly captured and was considerably enhanced under visible light irradiation with no DMPO-•O2- signal detected. This result indicates that •OH radicals are the main active species during the photocatalytic reaction. Fig. 7(a) shows that the DMPO-•OH signal of Bi4O5Br2 was stronger than that of Bi12O17Br2 under the same conditions, which suggests that the ability of Bi4O5Br2 to generate •OH radicals was stronger than that of Bi12O17Br2.
As shown in Table 1, the valence bands positions of Bi12O17Br2 and Bi4O5Br2were 3.10 and 3.29 eV, respectively; hence, the hole oxidation potentials of the valence band were higher than those of OH-/•OH (1.99 eV) and H2O/•OH (2.37 eV) [27, 28]. Thus, the hole oxidation potential of the valence bands could directly oxidize OH-(Eq. (5)) or H2O (Eq. (6)) to •OH radicals. In addition, the valence band potential energies of Bi12O17Br2 and Bi4O5Br2 were higher than those of EφNO2/NO (1.03 eV vs NHE), EφNO2/NO (0.99 eV vs NHE), and EφHNO3/NO (0.94 eV vs NHE); hence, the photo-generated holes of Bi12O17Br2 and Bi4O5Br2 could oxidize NO to NO2, HNO2, and HNO3 (Eq. (7)) [29, 30]. Furthermore, Bi4O5Br2 has a stronger valence band hole oxidation capacity than that of Bi12O17Br2 [26].
According to the electron paramagnetic resonance (EPR), paramagnetic defects are mainly generated by electron oxygen vacancies. The recombination of strongly bound holes and weakly bound electrons at oxygen vacancy centers can produce single ion oxygen vacancies [31, 32]. As seen from Fig. 8, the EPR signal of Bi4O5Br2 was considerably stronger than that of Bi12O17Br2, indicating that the oxygen vacancy content of Bi4O5Br2 was greater. The greater number of oxygen vacancies should favor bonding of O atoms to NO to promote the adsorption of NO by Bi4O5Br2.
Fig. 9(a) shows in situ DRIFTS spectra of Bi12O17Br2. The peaks at 1361 and 1425 cm-1 can be assigned to the NO2 [33, 34], which is the preliminary oxidation product of the photocatalytic reaction (Eq. (8)). At an earlier stage of the reaction, the NO2yield increased rapidly and then stabilized. The peak at 1249 cm-1 could be assigned to NO2- [33, 35], which became the strongest peak after 20 min of adsorption in the NO and O2 environment. There was no obvious change in the peak over the course of the photocatalytic reaction. Other peaks at 827, 880, and 1529 cm-1 were also attributed to NO2- [36, 37], which is a product of NO2 after receiving an electron (Eq. (9)). These three new peaks appeared during the irradiation process, and can be regarded as a new vibrational mode of NO2- [35]. The NO2- accumulated on the surface of photocatalyst leading to an increase of the peaks. The NO3- peak positions were located in the range 930–1030 and at 1777 cm-1 [38, 39]. The NO2- was further oxidized to NO3- by •OH radicals (Eq. (10)), and this became the final product of the system. Peaks in the range of 930–1030 cm-1 could not be detected and the peak intensity at 1777 cm-1 was also weak, which indicated that the yield of nitrate was comparatively low during the reaction process. Furthermore, the oxidation capacity of the Bi12O17Br2 sample was relatively weak, which is consistent with the results of the photocatalytic activity test.
The in situ DRIFTS spectrum of Bi4O5Br2 is shown in Fig. 9(b). The peak at 1072 cm-1 can be assigned to NO [34]. The larger specific surface area of Bi4O5Br2 enabled it to adsorb more NO to its surface during the 20-min adsorption process. As the illumination time was extended, the amount of adsorbed NO gradually increased and then tended to a dynamic equilibrium. The peak at 1638 cm-1 derives from NO2, and was rapidly consumed after the light was turned on, owing to combination with NO to generate N2O3 (Eq. (12)) with a peak at 1555 cm-1 [34-37]. The N2O3 gradually accumulated and was then oxidized to NO2- and NO3- by •OH radicals (Eq. (13)). The peaks at 818 and 1394 cm-1 could be assigned to NO2- [36-39]. The peak at 1394 cm-1 initially increased and then decreased, which indicated that NO2- was further oxidized by •OH radicals to form NO3- (Eq. (10)), with a peak located at 1273 cm-1. The peak of NO3- gradually became stronger and sharper, which reflected the high yield of NO3- in the photocatalytic reaction process. These results indicated that the Bi4O5Br2 sample had stronger oxidizing properties than the Bi12O17Br2 sample.
Compared with the Bi12O17Br2sample, the Bi4O5Br2 sample produced an intermediate product N2O3 during the photocatalytic process, indicating that the reaction pathways were different (Fig. 10). As suggested by the ESR results, the Bi4O5Br2 sample produced more •OH radicals, which enhanced the oxidation ability during the photocatalytic process. The presence of an intermediate product N2O3 reduced the reaction activation energy for the transformation of NO to NO3-, which in turn increased the yield of the final product NO3- [40]. consistent with the results of the photocatalytic activity test. In addition, the EPR results of the Bi4O5Br2 sample showed that more oxygen vacancies existed on the surface of Bi4O5Br2, which provided more reactive sites for the photocatalytic reaction to proceed at a constant rate. There was no obvious NO absorption peak for Bi12O17Br2, which explains its decline in photocatalytic activity over time.
In this work, Bi12O17Br2 and Bi4O5Br2 nanosheets with good visible light absorption properties were prepared by a simple hydrothermal treatment and room temperature deposition method, respectively. The Bi4O5Br2 nanosheets exhibited better photocatalytic activity and stability than those features of the Bi12O17Br2 nanosheets. Possible reasons for these different features are as follows: (1) Bi4O5Br2 had a higher specific surface area and larger pore size, which promoted the adsorption of the substrate and rapid transport of the reactants and intermediates, providing more active sites for the photocatalytic reaction; (2) Bi4O5Br2 nanosheets generated more •OH radicals and more valence holes formed during the photocatalytic reaction process with a stronger oxidizing ability; (3) O atoms in NO inserted into the Bi4O5Br2 oxygen vacancies, to provide more reactive sites; (4) The presence of intermediate N2O3 reduced the activation energy for the transformation of NO to NO3-. The present work shows that Bi4O5Br2 is a material with high visible light photocatalytic activity and good photochemical stability, which thus presents good potential for applications in environmental purification.
2017,
Vol. 38 Issue (12): 2030-2038 
PDF
2017,
Vol. 38 Issue (12): 2030-2038 PDF